← Back to community benchmarks

Qwen2.5-0.5B-Instruct

M3 (10c) · 16 GB · unknown · 2026-03-11

Performance

4k

tokens

2,479

PP tok/s

68.9

TG tok/s

1652

TTFT (ms)

1.5

Peak mem (GB)

Hardware

Chip M3 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.2.7

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	2,008	70.1	1.3 GB	view
4k	2,479	68.9	1.5 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	70.1	1.00×
2×	81.7	1.17×
4×	184.8	2.64×